INDEX

Explanations

phrases related to control or authority over a situation or entity

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

occ

-0.67

icut

-0.66

onge

-0.64

 subp

-0.61

Ann

-0.60

autical

-0.58

etheless

-0.58

rely

-0.57

anus

-0.56

 Scotia

-0.56

POSITIVE LOGITS

FUL

0.74

ulator

0.65

maid

0.61

 corrupted

0.61

lopp

0.60

 wielded

0.60

offs

0.56

books

0.55

idable

0.55

 corrupt

0.55

Activations Density 6.603%

phrases related to control or authority over a situation or entity

No Comments

No Known Activations